Dissertation > Excellent graduate degree dissertation topics show

Image Retrieval Based on Color and Non-Subsampled Contourilet Features

Author: ZhangHuiYun
Tutor: ZhangXinMing
School: Henan Normal
Course: Applied Computer Technology
Keywords: Image Retrieval Significant regional Histogram Nonsubsampled Contourlet Transform Information entropy
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 24
Quote: 0
Read: Download Dissertation

Abstract


With the rapid development of information technology, vast amounts of digital images and videos continue to emerge. Content-based image retrieval technology came into being in order to effectively use these resources, and quickly became a hot research direction in the field. This technology is primarily the underlying characteristics of the extracted image, and its similarity measure. How to extract and use the bottom of the image characteristics, making the retrieval to quickly and effectively is the key technology of the research in this field, is currently the main issues that need to be addressed. The image of the underlying characteristics including color, texture, shape and spatial relationships, this paper focuses on color and texture features of the image-depth analysis and research, combined with the image color histogram and nonsubsampled Contourlet transform theory, put forward a comprehensive image significantly regional histogram and nonsubsampled Contourlet domain texture feature retrieval methods. The main research work and innovations are as follows: 1, the detection of a significant point of a significant regional the positioning image of ring, and then were extracted from the image significantly histogram of the area and the background area, a combination of the two histograms weighted method, this significant regional color histogram method. This approach not only be flexible enough to obtain the image a significant area makes the color information having spatial, but also to ensure the rotation of the image unchanged. This method overcomes the shortcomings of the currently used method of extracting color features can effectively extract color features significantly improve the image retrieval results. 2, through the study of multi-resolution nonsubsampled Contourlet transform theory, proposed a nonsubsampled the Contourlet information entropy texture description method. The method is obtained in the decomposition of the image when a strong direction, translation-invariant, but also effectively reduces the number of dimensions of the texture features, has certain advantages for image retrieval. 3, the color and texture of the proposed method for fusion, presents a comprehensive significant regional histograms and nonsubsampled Contourlet texture characteristics of image retrieval methods. The combination of these two features, respectively, will be described from the global and local features of the image, the complementary mutual defects. The paper design a comprehensive color and texture features in image retrieval model, to build retrieval system using this model, and the above mentioned methods are realized, and use various methods to carry out the experiment. Experimental results show that the stability of this paper, this method performance has a higher retrieval efficiency than single feature and traditional features.

Related Dissertations

  1. Quantitative evaluation model based on information entropy of the classroom observation,G632.4
  2. Research of High Speed Image Pre-processing System Based on FPGA,TP391.41
  3. Application of Q-Learning in the Content-Based Image Retrieval Technology,TP391.41
  4. Research and Implementation on Content-Based Clothing Image Retrieval,TP391.41
  5. The Discovery of User Concept Region Based on Multiple Instance Learning,TP391.41
  6. Research and Application of Diverse Density Learning Algorithm,TP181
  7. Based on statistical methods of magnetic resonance imaging of the human brain image segmentation and three-dimensional analysis of the data,R445.2
  8. Research of Color Correction Method for Multiview Video,TP391.41
  9. Capsule endoscopy and endoscopic image portable receiver system bleeding Recognition Algorithm,TP391.41
  10. Content-based image retrieval technology research large-scale digital,TP391.41
  11. Histogram based on improved gradient space object detection method,TP391.41
  12. Based on local binary pattern and Weber local descriptor Face Recognition,TP391.41
  13. Feature points based spatial information distribution histogram matching method,TP391.41
  14. Exam cheating Intelligent Analysis Method,TP391.41
  15. Statistical model based on adaptive histogram multi-threshold image segmentation algorithm,TP391.41
  16. Infrared image enhancement and color lookup table -based technology research,TP391.41
  17. Research of Orthodontics Image Retrieval Based on Multi-Feature,TP391.41
  18. Research on Method of Integrated Multi-Features Based Image Retrieval,TP391.41
  19. Application of Drilling Information Analysis and Supervise by MAPGIS,P208
  20. Remote Sensing Image Retrieval Based on Radiation and Spatial Information,TP751
  21. The Study of Application of Cardiac Troponin T Combined with Dose-volume Histogram (DVH) in Detecting Radioactive Myocardial Damage,R818

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile